This paper presents the preliminary results of a study on the acquisition of Indonesian by an Indonesian-Italian bilingual child, focussing on negative constructions. It is based on naturalistic data collected in Jakarta from Guglielmo, a child born to an Italian mother and an Indonesian father, and living in a predominantly Indonesian-speaking environment. Guglielmo has been video-recorded from the age of 8 months on a weekly basis; at the time of writing his age was 2;0. The purpose of the study is to examine whether the two languages are separated in Guglielmo's mind, and when this separation occurs, or-if there are two coexisting systems-to what extent they influence each other.

Guglielmo has a passive competence of Italian, and is able to understand most utterances addressed to him by his mother. His knowledge of Indonesian, on the other hand, is fully active. An analysis of the data indicates that at this stage the two are functionally merged. Guglielmo does not alter his speech when addressing his Italian mother on the one hand, or his Indonesian father (and other Indonesian interlocutors) on the other hand. He uses a single code, basically Indonesian with a few Italian elements, with all the interlocutors.

Most of Guglielmo's vocabulary is Indonesian, but it contains some Italian lexical items as well. Interestingly, when producing Italian words, he produces geminate consonants correctly, even though gemination does not occur in Indonesian. Words produced with geminate consonants include mamma 'mother', nonna 'grandmother', latte 'milk', and dillo (<coccodrillo) 'crocodile'. It is interesting to note that there are no words in Guglielmo's active vocabulary which occur in both Indonesian and Italian versions; words in the lexicon are consistently either Indonesian or Italian.

The production of negative sentences where Italian and Indonesian elements are mixed supports the hypothesis of early bilingual language merger, as exemplified in sentence (1) below:

(1)	Mimik	latte	no,	Ma
	drink	milk	NEG	Mom
	'I don't want to drink milk, Mom.'
	(while giving back the bottle to his mother)

In Italian, there are two negative particles. Non is used to negate verbs, nouns, pronouns, and adjectives, while no is used to negate entire propositions. In Jakarta Indonesian, three different devices are used for negation. Generally speaking, nggak negates verbs and adjectives, bukan negates nouns and entire propositions, and jangan is used in prohibitions.

Guglielmo distinguishes between three different functions of negation: rejection or refusal of a request or an action; the non-existence or absence of something; and the denial of the truth or accuracy of a statement. For general negation, he consistently uses 'no', by attaching it to the end of affirmative sentence like (1) above. For non-existence of something, he uses tak/ndak ada (lit. 'there isn't') or tak/ndak bisa (lit. 'cannot') that basically can be interpreted as a single unit, as he never uses 'tak/ndak' separately or with other items. Finally, bukan (pronounced [utan]) has been used consistently since the age of 1;8 only to mark an opposition. It has an emphatic intonation and generally anticipates an implicit contradictory clause.

(2)	Bukan	ini,	Ma!
	NEG	this	Mom
	'Not this one, Mom!'
	(rejecting an offer to watch a film on TV, but willing to watch another one)

(3)	Bukan,	mimik	Gulli	ini!
	NEG	drink	Guglielmo	this
	'Not (that one), this is my drink'
	(refusing to drink milk from a certain bottle and indicating another one instead).

Finally, two negative particles, common in adult Jakarta Indonesian, have not been acquired yet by Guglielmo. Belum 'not yet', used to give a negative answer to a question marked by the perfective sudah, never occurs in Guglielmo's speech. In response to questions like Sudah makan? 'Have you eaten yet?', he answers affirmatively with a simple iya 'yes' or uda 'perfective particle' (as adults would), or negatively with no. Finally, Guglielmo has yet to acquire jangan, used by adults to form negative imperatives. Instead, he also uses no.

The results of this study are supportive of a hypothesis of a mixed system. However, when evaluating them, other factors like social and contextual influences also have to be taken into account, to explain whether these social forces might have an impact upon the child in his language learning.